On Friday evening the much anticipated WolframAlpha launched publicly.  It is great fun to play around with, not only because it has some hilarious easter eggs, but because there is a renewed sense of exploration.  At least I found myself coming up with ever new queries to see what it knows (computes) and what it doesn’t.  For instance, here are two queries that show off the range of WolframAlpha:

• how old is Barack Obama?
  

• area under x^2 between 2 and 4?

One thing is for sure – a lot of homework will never be the same!  But what about beyond that?  After some pre-launch coverage calling it a potential “Google Killer” and google’s less-than-subtle trumpeting of its own structured data efforts there has been an unreasonable amount of expectation built up and not surprisingly some of the initial reviews are now lukewarm.

I think that a lot of what happens next depends on how quickly WolframAlpha can open up their system for participation by third parties.  The power of a web search engine derives from the fact that the underlying corpus is constantly growing through many individual additions.  The additional structure required for computational search meant that WolframAlpha had much of its corpus manually prepared in a controlled fashion.  That of course does not scale.  If WolframAlpha can open up the system to third party contributions of data and eventually of algorithms, then it will take us a big step closer to that omniscient computer from the Star Trek bridge.

In the meantime, it is nice to see something new and differentiated in the market place.  As Dare Obasanjo points out, WolframAlpha may fit with a pattern of using different search engines for different purposes.  And if nothing else, this will make sure that Google gets better faster!